version 7.0log using JVREPcorrectdata.log, replace#delimit ;*     ***************************************************************** *;*     ***************************************************************** *;*   File-Name:  JVREPcorrectdata.do                                     *;*   Date:       June 12, 2001                                           *;*   Author:     MRG                                                     *;*   Purpose:    Replication of Jackman and Volpert's 1996 BJPS piece    *;*               on far right parties using the correct data.            *;*   Input File: JV.dta                                                  *;*   Output File:JVREPcorrectdata.log                                    *;*   Data Output:JVREPcorrectdata.dta                                    *;*   Previous file:                                                      *;*   Machine:    system7                                                 *;*     ****************************************************************  *;*     ****************************************************************  *;use h:\JackmanReplication\replication\JV.dta;summarize; rename countr country;rename elyear year;label var year "election year";label var erps "extreme right percentage support";rename unem unemp;label var unemp "unemployment rate";rename state1 austria;rename state2 belgium;rename state3 denmark;rename state4 finland;rename state5 france;rename state6 germany;rename state7 greece; rename state8 ireland;rename state9 italy;rename state10 netherlands;rename state11 norway;rename state12 portugal;rename state13 spain;rename state14 sweden;rename state15 switzerland;rename state16 unitedkingdom;rename enpp enp;label var enp "effective number of parties";label var lerps "natural log of erps + 1";label var thresh "effective threshold";rename thresxpp threshenp;label var threshenp "effective threshold x enp";*     ****************************************************************  *;*       Jackman and Volpert contain some data entry errors. They missed *;*       the 1971 Danish election and added elections for Ireland in     *;*       1990 and the UK in 1989 that did not take place. I drop the two *;*       elections that did not happen and add the Danish election       *;*       with data from Mackie and Rose.  Jackman and Volpert also       *;*       used incorrect figures for the effective number of parties      *;*       and effective threshold in the 1970 election in Austria         *;*       (correct figures come from Lijphart i.e. the same source that   *;*       Jackman/Volpert cite).  These need to be corrected for.         *;*     ****************************************************************  *;drop in 103;drop in 55;replace enp = 2.25 in 1;replace thresh = 8.5 in 1;set obs 102;replace country = "denm" in 102;replace year = 1971 in 102;replace unemp = 0.9 in 102;replace austria = 0 in 102;replace belgium = 0 in 102;replace denmark = 1 in 102;replace finland = 0 in 102; replace france = 0 in 102;replace germany = 0 in 102;replace greece = 0 in 102;replace ireland = 0 in 102;replace italy = 0 in 102;replace netherlands = 0 in 102;replace norway = 0 in 102;replace portugal = 0 in 102;replace spain = 0 in 102;replace sweden = 0 in 102;replace switzerland = 0 in 102;replace unitedkingdom = 0 in 102;replace enp = 4.92 in 102;replace lerps = 0 in 102;replace erps = 0 in 102;replace thresh = 2 in 102; replace threshenp = 9.84 in 102;sort country year;summarize erps lerps thresh enp unemp; *     ****************************************************************  *;*       Note that there are slight changes to the summary statistics,   *;*       but nothing too large.                                          *;*     ****************************************************************  *;*     ****************************************************************  *;*       Do these changes in the data affect the results?                *;*       Yes, look at model 2 which is JV's main model.  Here ENP is no  *;*       longer significant at the 95% level.                            *; *     ****************************************************************  *;*     ****************************************************************  *;*       Replicate Model 1: page 514.  Tobit with no interaction terms.  *;*     ****************************************************************  *;tobit lerps thresh enp unemp austria belgium denmark finland france    germany greece ireland italy netherlands norway portugal spain     sweden switzerland, ll(0) ltolerance (1e-5);    *     ****************************************************************  *;*       Replicate Model 2: page 514.  Tobit with interaction terms.     *;*     ****************************************************************  *;tobit lerps thresh threshenp enp unemp austria belgium denmark finland france    germany greece ireland italy netherlands norway portugal spain     sweden switzerland, ll(0) ltolerance (1e-5);*     ****************************************************************  *;*       Replicate Model 3: page 514.  OLS with interaction terms.       *;*     ****************************************************************  *;regress  lerps thresh enp  threshenp unemp austria belgium denmark finland france    germany greece ireland italy netherlands norway portugal spain     sweden switzerland if lerps>0;                 save    JVREPcorrectdata, replace;exit;